Recognizing sea animals is a significant challenge in digital image recognition. This is due to the diverse visual characteristics of marine animals, including morphological shapes, body surface colors, and textures displayed in images. Environmental factors also influence image quality, such as underwater lighting conditions, water turbidity, and other external elements. To address these classification challenges, one proposed approach is the use of the Extreme Learning Machine (ELM) method, which can be implemented by utilizing HSV (Hue, Saturation, Value) color features as the main input. The HSV color space is chosen because it more closely resembles the way humans perceive colors. In this model, color is separated into three main components: hue represents the type of color, saturation indicates the intensity or purity of the color, and value refers to its brightness or darkness. The dataset consists of several classes of marine animals such as clams, squids, and shrimp, collected from high-resolution image datasets. Test results show that the ELM model can classify images with competitive accuracy, achieving up to 83% accuracy in a much shorter training time compared to traditional learning methods. This study demonstrates that combining HSV color features with the ELM algorithm can be an efficient approach for classifying marine animal images.
 
Keywords - Shell, Squid, Shrimp, ELM,HSV